The Coast Guard published a notice of intent in the Federal Register on 26 October saying it plans to enter a cooperative research and development agreement with LOOKOUT to assess the company's Super Intelligence tools for automated target detection, and to establish a coverage-factor value for data-driven Lateral Range Curve results in search and rescue.
The notice says the service currently draws on many sensor inputs across units and assets, and that detection tools could help it automatically detect, classify and track targets while handling low light, fog, precipitation and rough sea states. It frames the work as benchmarking, interoperability testing and proposals for data-handling practice, plus characterising system confidence and operational performance so that a measurable evaluation tool can lower the barrier to integrating such systems.
Proposed tools must be at a high technology-readiness level, commercially available, self-contained, portable, capable of edge computing and compatible with video-feed-agnostic sensors. The Coast Guard is taking comment on the proposed agreement until 25 November and says other federal entities with relevant capabilities may propose similar agreements.
Our takeThe technical question is what a coverage factor derived from a commercial detector actually measures. Lateral Range Curves assume a detection probability that varies with sweep width; plugging in a vendor figure without published performance across sea state, light level and sensor type only relocates the guesswork. The notice's own list, self-contained, portable, edge-computing and sensor-agnostic, is a demanding combination for a model that has to run on a boat. Worth watching whether the agreement yields a test protocol others can reuse rather than a single integration.